Michael E. Gadringer

Efficient Power Amplifier Identification Using Modified Parallel Cascade Hammerstein Models

The automatic generation of behavioral amplifier models based on the measurement of the input and output signals imply demanding requirements for the identification algorithm. In addition to the computational efficiency it is very important that such an algorithm needs only a low number of predefined parameters for the fitting process. To fulfil these requirements we suggest the use of a modified parallel cascade Hammerstein model (MPCHM). To verify the performance of the presented algorithm the behavior of a 3-stage RF power amplifier (PA) was identified. The resulting model was validated using different measurements with similar statistical properties. Furthermore the static nonlinear contributions were deembedded from both the measured and the modeled output signals to visualize the capability of the MPCHM in predicting the amplifier memory effects

Heuristic Algorithms for Power Amplifier Behavioral Modeling

This letter presents the use of two heuristic search algorithms, named simulated annealing and genetic algorithms, for the extraction of power amplifier (PA) behavioral model parameters. Their application in this letter consists in determining the memory length and the most significant delays of the considered model structure. Two PA behavioral models have been considered: an augmented nonlinear moving average model and a nonlinear auto-regressive moving average model. By using WCDMA signals measured from a three-stage LDMOS class AB PA, both PA models were extracted. Finally, results presenting the advantages of using these heuristic search algorithms are provided.

RF-power amplifier characteristics determination using parallel cascade Wiener models and pseudo-inverse techniques

This article presents a black-box modeling approach of a microwave power amplifier (PA) and defines the linear and nonlinear operation regions. An in-band quasi-white real-valued noise like signal is used as stimulus for the identification process to excite every possibly source of nonlinearity. A segment of the input-output measurement data is processed to generate an initial parallel cascade Wiener model (PCWM). The first order Volterra kernel is extracted in order to estimate the amplifiers memory. Then pseudo-inverse techniques are used to find a parsimonious model. This model is cross-validated with the entire measurement data. The presented modeling approach results in a model intended to be numerically robust and having a high identification percentage based on a variance figure of merit. Finally the linear and nonlinear operation regions of the amplifier are defined, as an alternative to the use of AM/AM curves.

Analysis of Variations of Volterra Series Models for RF Power Amplifiers

This article discusses several different implementations of Volterra series models, and performs a fair comparison between them using measured data and well known figures of merit. Among the implemented models, an approach that uses sub-sampling techniques together with Volterra series in a parallel cascade structure shows an improved performance compared to the other variations of Volterra series models. Based on the results, new trends in RF power amplifier behavioral modeling are suggested.

Characterization and Modeling of Direct Conversion Transmitters

This work presents a description of a direct conversion transmitter excluding the power amplifier. The model predicts the impact of the even- and odd-order distortion introduced by the analog baseband signal processing and the mixer together with linear effects. The model is parameterized by measuring the response of a direct conversion transmitter onto single- and two-tone signals. By comparing the response of the transmitter to the model output the performance of the presented approach is verified.

Loop controller for a feedforward amplifier minimizing the measured power signal

Adaptively controlled feedforward amplifiers are discussed in a large number of publications over the past years. Especially for the control of the distortion cancellation loop it was inevitable to implement a baseband-controller if the use of a pilot carrier should have been avoided. A significant deterioration of the costs of a feedforward amplifier can be achieved by using power sensors to balance all loops. This paper introduces a loop controller, which tunes the loops of a feedforward amplifier based on the power measurement of two signals within the circuit. The structure of this controller is discussed and measurement results for a 40 W GSM2+ (EDGE) power amplifier are presented.

Comparison of the imbalance effects in direct conversion transmitters and receivers

Although direct conversion transmitters and receivers share the same structure certain differences in their behavior can be found. These differences result in diverse performance limitations. This paper derives a linear model of direct conversion transmitters and receivers. Based on this model methods to extract the frequency dependent characteristics are introduced and compensators are suggested. By applying these techniques to a custom transmitter and receiver setup the achievable imbalance suppressions are demonstrated and the derived performance improvements are compared.

Design of a current mode class-D RF amplifier using load pull techniques

This work presents the design of a current mode class-D (CMCD) amplifier using a measurement based approach. Because the CMCD amplifier is in fact a push-pull version of the inverted class-F amplifier the optimum load impedances for an inverted class-F single-ended operated LDMOS FETs have been determined using an active harmonic load-pull setup. The correct time domain waveforms for an inverted class-F amplifier were verified by measurements and simulations. Based on the harmonic load-pull measurements an experimental current mode class-D amplifier with 61.5% power-added efficiency (PAE) at 25W output is demonstrated using LDMOS FETs at 900 MHz.

Analysis of RF-Power Amplifier Modeling Performance using a 16-QAM Modulation over AWGN Channels

This paper presents the black-box modeling of a microwave power amplifier (PA) and its performance in the nonlinear operation region - low input back off (IBO). As input signal a rectangle 16-QAM corrupted by different levels of additive white Gaussian noise (AWGN) was used. The nonlinearities caused by the amplifier in low IBO operating conditions can change the output constellation and lead to difficulties in finding the correct model parameters, as the black-box identification procedure is based on measured input/output signals. These signals can be altered by different types of noise. The AWGN degrades the signal to noise ratio (SNR) and can lower the model accuracy that should ideally not be influenced. In this paper the models were first estimated using noise-free data (NFD) and secondly with different levels of noisy corrupted data (NCD). NFD was also applied to the models calculated with NCD and vice versa. An investigation of the loss of accuracy caused by the noise in the identified models is presented. Three situations were considered: 6dB IBO, 4dB IBO, and 2dB IBO. It is shown that the pseudo-inverse based models present reasonable results for PA modeling when AWGN is incident

Phase Dependent Distortion in Direct Conversion Transmitters

The impairments and nonlinear (NL) distortion generated by the analog quadrature modulator in a direct conversion transmitter affect the performance of a transmission system. These effects cause input signal dependent gain variations at the low-pass equivalent I/Q-modulator output. The aim of this paper is to analytically derive the relationship between the mean transmitter gain and the tone phases of single- or two-tone excitation signals. This relationship is separately evaluated for the different distortion sources in the direct conversion transmitter. It will be shown that only two-tone signals with frequencies symmetrical to DC are able to excite these phase-dependent gain variations. Additionally, the conditions for the presence of the distortion mechanisms are highlighted. Using measurements of an experimental transmitter setup, the validity of the derived results is shown. The presence of phase-dependent gain changes were also measured when applying a DC-symmetric multi-tone signal.